This proposal aims to elucidate the contributions of medial-temporal lobe (MTL) and frontal-lobe regions to the encoding and retrieval of declarative memory via functional magnetic resonance imaging (fMRI). Dysfunction in MTL and frontal regions is thought to be important for many neurological and psychiatric diseases, including Alzheimer's disease, schizophrenia, depression, epilepsy, and Parkinson's disease. There are, however, major contradictions in the functional neuroimaging literature in characterizing MTL and frontal-lobe contributions to declarative memories. The overall goal of the studies is to develop a reliable understanding of the meaning of MTL and frontal activations during declarative memory performance, so that neuroimaging studies may become better integrated with evidence from animal, human lesion, computional, and theoretical approaches. Experiments 1-5 exploit recent advances in the use of event-related fMRI to discover the anatomic locus of activations, within the MTL, correlated with sucessful encoding or retrieval memories for scenes (Experiment 1), words (Experiment 2), faces (Experiment 3), nonverbal patterns (Experiment 4), and sentences (Experiment 5). Experiment 6 examines different ways of measuring encoding. Experiments 14-15 focus on differences across materials. The critical questions are similarities and differences for (a) encoding vs. retrieval and (b)different stimulus types. Both lesion and neuroimaging evidence suggests that the amygdala enhances declarative memory for emotional, especially negative, visual scenes. In Experiments 7-8 we examine MTL activations for negative (Experiment 7) or positive (Experiment 8) vs. neutral scenes. We hypothesize that frontal-lobe activations commomly observed during declarative memory encoding and retrieval reflect domain-specific and process-specific working memory processes interacting with long-term memory stores. Experiments 9-10 test the hypothesis that frontal-lobe retrieval activations are co-localized for short-term and long-term memory judgements. Experiment 11 tests the hypothesis that left-prefrontal encoding activations reflect working-memory processes of verbal selection among competing alternative responses. Experiment 12 tests the hypothesis that there is material-specificity for frontal-lobe encoding and retrieval processes. Experiment 13 examines whether the material-specific hypothesis applies to both task-based and material-based operationalizations of encoding. Experiment 16 examines what frontal-lobe component of working memory (encoding, storage, or retrieval) participates in declarative memory retrieval.